Vodia PBX
-
Posts
11,111 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Posts posted by Vodia PBX
-
-
Logfile shows "Set processor affinity to 1 failed" on each service startup.
How do I fix this?
First, it should also be possible to lock the processor to a core from outside of the program itself. Seems http://www.cyberciti.biz/tips/setting-proc...or-process.html is a interesting link to do that.
Maybe the apt-get also installs the neccessary stuff so that the processor can do it on its own. I think the link above is interesting reading regarding this topic.
-
ok i figured that would get attention
But when I pitch a snom phone and show them the record button and have to say to enable that button you need to spend double.. yea i loose out.
So pbxnsip needs to enable ad-hoc recording at part of the base licenses. Full time recording should be its own separate license.
then it would be easier to sell snom phones.
Not all countries allow recording of calls. So we definitevely have to have a license key for recording, even user-initiated.
-
For those who can live without extensive release notes, there are new images for 3.1 out:
http://pbxnsip.com/download/pbxctrl-3.1.0.3043.exe (Windows, manual upgrade)
http://pbxnsip.com/download/pbx3.1.0.3043.exe (full Windows installer)
http://pbxnsip.com/download/pbxctrl-freebsd7.0-3.1.0.3043 (FreeBSD7.0 32-bit)
http://pbxnsip.com/download/pbxctrl-rhes4-3.1.0.3043 (RedHat 4)
http://pbxnsip.com/download/pbxctrl-centos5-3.1.0.3043 (CentOS 5)
http://pbxnsip.com/download/pbxctrl-suse10-3.1.0.3043 (SuSE10)
http://pbxnsip.com/download/pbxctrl-debian4.0-3.1.0.3043 (Debian 4)
http://pbxnsip.com/cs410/update-3.1.0.3043.tgz (CS410)
We'll have the release notes and the missing OS (e.g. MacOS) ready probably on Tuesday, until then please drive carefully!
-
Pre auto-atendant greeting, lets says good morning, or good afternoon, or good evening, etc. or we are currently closed, then play standard greeting.
If the service flag related to *9870*1 And *9870*2 and *9800*3 are all active play all 3 greetings, not just the first.
*9870*1 can be good morning
*9870*2 can be we are currently closed
*9800*3 can be auto attendant greeting
:-S the workaround at this point is to record always the following message together with the morning/afternoon/happy Xmas...
-
I played with the IVR Tab quite a few times. Although you can set several greetings according to a service flag, it will only play 1 greeting, if two service flags are active it will only play the first, the standard greeting will only play when no service flag is active. So a welcome greeting can not be activated.
Oh so you want to have a pre-welcome greeting?
-
I also think a pre-autoattendant greeting would be great. The way I have it now is 3 complete recordings with the same message, just the first 2 words change, good morning, good afternoon and good evening, and I have a service flag scheduled for the hours per day.
Right now if you activate 2 greetings with the service flags, only the first plays, you can leave the functionality as is, just allow activating multiple greetings.
so if I have 3 messages.
Good morning set
We are currently closed set
You have reached the autoattendant, if you know your party... set
It should play Good morning We are currently closed You have reached the autoattendant, if you know your party...
Right now it would only play good morning.
Maybe I don't understand this... There is already a IVR tab for the auto attendant where you can specify depending on service flags what welcome greeting should be played back.
-
-
Does anyone know how to transfer a call straight to voicemail if the user is already on a call. One of our clients has asked and don't see it in the feature list in the domain settings. Thanks for the help.
P.S. If this isn't possible we really could use this... Also would like to have a section that we can post a wish list of sort.
Well, you can set the "lines" parameter of the extension to "1". That means there is only one call going to that extension at a time.
-
Client said that the calls cannot be dropped so they are increasing the call duration to infinite. Is there any way to have the calls detected and dropped?
Well, that's what the settings "timeout_conference" is good for. If a user-agent "sends" one-way audio (which is no audio) that is okay. There are many ways around it, for example choosing a user-agent that sends keep-alive RTP (silence indicators). But we want to compatible to devices that are not so smart.
I would not change the general maximum durtion of the call. It is a uneccessary burden to the system.
-
No it does not process any futher than that.
The Wireshark just sits there.
That is why I suspected that PBXNSIP was nor responding to the "Response: 500 Syntax error, command unrecognized" message.
Hmm. I remember there was a problem with those old servers. Might be fixed already. Maybe you can wait a few days and then try the upcoming 3.1.
-
Here is are the three lines of Wireshark I receive when trying to use one particulat Email Server:
Line_____Frame______Originate IP_____Destination IP__Service
2479___13.643247___211.146.45.24___34.41.150.162___SMTP___Response: 220 APPSERVER1.XXX.local SMTP service ready
2481___13.644240___34.41.150.162___211.146.45.24___SMTP___Command: EHLO localhost
2483___13.659967___211.146.45.24___34.41.150.162___SMTP___Response: 500 Syntax error, command unrecognized
It looks like the Email Server does not understand the newer EHLO command.
Under this condition the Email Client would then send the older HELO command.
PBXNSIP does not send the HELLO command.
Is this a bug or a missing condition?
Well that message itself is not a problem. It is actually the only way to find out if the server understands EHLO.
Does the rest of the email delivery work?
-
I have found in 3.1.0.3040 if you disable a mailbox, it still answers, and you can still leave, and retreive a message.
Ops, good catch! 3.1 is hopefully becoming a reality very soon.......
-
421 693.644257 205.240.200.62 192.168.1.3 SIP Request: BYE sip:5045403107@192.168.1.3:5060
422 693.647387 192.168.1.3 205.240.200.62 SIP Status: 200 Ok
Well that looks like your service provider disconnects the call... It is really strange after 11 minutes. Maybe they have a 11-minute demo key from their softswitch vendor ...
No it seems that the re-INVITE fails. Looking into it... Can you send us a private mail with your account information so we can give it a try from here?
-
Where is this "hidden" setting?
Sounds like there is nothing to worry about then.
This will make for a happy client!
Thanks!
Well, it is a global setting (see http://wiki.pbxnsip.com/index.php/Global_Configuration_File). There is also a small trick to set the setting through the web interface. It is not visible through ther web pages because we usually don't want people play with these settings.............
-
Next.. in this process
The call only lasts 11 minutes with 11 seconds
I check the value of "Maximum call duration" and it is 7200
Who disconnects the call - the service provider or the PBX? (Does the PBX send the BYE request or the "200 Ok" response?)
-
I have a client using that has a unique application. They are using a push button/speaker/mic device to automatically connect to pbxnsip and connect to a conference room. That works fine. All the devices are in the same conference room and this lets everyone communicate like an intercom system.
If the devices reboot there is no bye packet being sent. When they boot back up and automatically dial the conference room again the call status shows that the same extension now has 2 calls in there.
Will this cause problems after a couple days or weeks or months?
The extension is only showing 1 registration since it is from the same IP. They are just concerned with the extra, no longer connected calls. Will this eventually crash the system as the CPU starts going up and up from all these simulaneous calls?
I don't think it matters but they are using the 410 and 425 appliances but it also does this on windows and linux. They have also tried 3CX and had the same results so it's not PBXnSIP specific.
What can be done to prevent problems or is this a non-issue?
After the maximum call duration, the call gets disconnected anyway. The default is two hours.
You can make this setting shorter if you like; though I would believe it is a resonable setting. There is also a "hidden" setting called "timeout_conference" which is by default 3600 seconds. This value was so high because there are user-agents out there that do not send keep-alive traffic when mute-ing a call. If you have devices that do send keep-alive then it is safe to lower this value to something like 600 (10 minutes).
-
I have an AASTRA 57i CT that has been restored to defaults that I am trying to get to work with my PBXNSIP with no luck.
The only modifications I have done thus far is under the SIP Settings> Proxy IP/Port > Proxy Server, I put the address of the PBXNSIP.
Can anyone tell me what else needs to take place or happen to make this phone register to an extension?
This is an in-house demo so security is no factor here. I am looking for the easiest route in making this phone make and receive calls to other extensions.
Actually, plug and play should work. Punch in the IP address into the "TFTP server" setting into the phone, and set the Admin->Settings->Ports->TFTP policy to always send the password and set at least one extensions MAC address to "*" and give it a try.
Otherwise, just make sure that you fill in the outbound proxy on the phone and also use a domain name in the line settings.
-
We are using Addressbook in PBXnSIP extensively for gatehouses to dial a stand number, which in turn phones the client. Theproblem is that the CDR created shows the speed dial code rather than the actual number phones, which means the gatehouse making the calls cannot be billed accurately (or at all) as the dialled number in the CDR shows *1234 rather than +27 12 661 1177. This is a major issue as billing is a nightmare to compare the CDR's to addressbook to determine the number that was dialled. Urgent assistance required.
What version are you on? Could be that we already changed that in the meantime.
-
isnt there a way to just see SIP traffic? in real time? Invite messages and such?
I can see them through an edgemark but the must have a modified version of tcpdump.
Well, the SIP traffic can be seen through the web interface of the PBX (well, filtered out keep-alive stuff). The tcpdump is a tiny version, but maybe it does have the port option.
-
yes. but what happens if the PBX only has a local IP (in the DMZ scenerio) and I want to have a couple offsite phones. Right now I manually config the phones. but if they could be created via PNP TFTP that would save time.
Most everyone else cant do this, but in my mind it seems simple thing to do.
You can still override the OS routing mechanisms, and that also applies to stuff like TFTP (see http://wiki.pbxnsip.com/index.php/Office_w...ic_IP_addresses). A typical case is DMZ.
And, of course, DMZ is a bad thing to do in SIP environment. It is a lot easier to have a routable IP address, a public IPv4 or IPv6 address.
-
I used to use "tcpdump -n -s 0 -i any port 5060 " and i could see sip messages.. now i cant.
But that doesnt seem to work on the cs410.. what command should i use?
tcpdump is the right command. Not sure about the "-i any", maybe you have to specify what interface you want to listen on.
-
yes however how would the PNP system know to use the internal number 10.1.1.1 or the external number 208.208.208.208 as the proxy? have to be some sort of internal/external toggle in the system.
The source if the TFTP address can tell the PBX what interface to use. BTW there is nothing like an "internal" or "external" address. Think about a system that has ten IP addresses (LAN, WAN, VPN1, VPN2, WiFi, IPv6, ...). The OS routing table is responsible for determining what IP address to use.
-
It would be cool you could have a setting to tell the TFTP system if the phone was internal or external. so the PnP system would use the appropriate IP.
It should do that (of course). It should be possible to provision internal and external phones through the same TFTP socket.
-
TLS shouldn't start until the RTP starts right? hopefully DNS is cached.. maybe its hard for the system to generate so much with internal database lookups but i would think the bandwidth should be fairly light with no RTP.
No, TLS is used to send out the INVITE through a secure channel. In SIP, you have to allocate the RTP ports when sending out the INVITE (unless you want to end up in an endless discussion about offer/answer). So it is a pretty expensive operation.
Apart from that, I think if all 50 phones are ringing when someone calls in, not sure if that is what you want in an office.......
select returns error 22
in Linux-related Topics
Posted
Errno 22 is "Invalid Argument" - seems like the timeout for select() got screwed up in that case. "Should not happen" - but it seems we need to put an extra safety belt on here.